Hyperglycemia and insulin therapy in the critically ill child

To canvass the current opinion of pediatric intensivists in the United Kingdom regarding the importance of hyperglycemia,their approach to management, and their views in relation to a potential intervention trial of tight glycemic control.

Electronic survey comprising a 17-point questionnaire,along with six clinical scenarios describing cases of bronchiolitis, septic shock, major trauma, postcardiac surgery, necrotizing enterocolitis,and acute respiratory distress syndrome.

All pediatric intensive care units in the United Kingdom(n = 21).

All consultant pediatric intensivists in the United Kingdom (n = 117).

None.

The response rate was 75% (88 of 117), representing all 21 units. Although all respondents administer insulin to hyperglycemic patients, only one in two believe hyperglycemia causes significant morbidity in critically ill children. Predominant factors influencing the decision to start insulin are severity of hyperglycemia (95% of responders), duration of hyperglycemia (85%), and patient diagnosis (49%). The blood glucose threshold for starting insulin varied widely from 6.1 to>15 mmol x L(-1) (110-270 mg x dL(-1)) for each of the six scenarios,with poor agreement between intensivists within each center(intraclass correlation coefficient of 0.29). Sixty-seven percent of respondents, representing 14 of the 21 units reported having no written policy or guideline in their unit for management of hyper-glycemia in nondiabetic children. Eighty-six percent (76) of intensivists would be prepared to participate in a trial of "tight glycemic control" but only 63% (55) would be prepared to include all patient groups. Only 20% would be willing to target blood glucose below 6.1 mmol x L(-1) (110 mg x dL(-1)) in the tight control group.

This survey suggests significant variation in the management of hyperglycemia across the UK Practice varies even among intensivists from the same unit, reflecting the fact that few units have 'an agreed written guidance in place.' The majority of intensivists would be prepared to participate in a trial of tight glycemic control.